{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Replication Data for \"The Limited Influence of Right-Wing Movements on Social Media User Engagement\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Disclaimer: This notebook includes the Python code used for creating tables and figures for main paper as well as the appendix. Unfortunately, I am not allowed to publicly share the data files required for replication. In case you would like to get access  please reach out to me at c.schwem2er@gmail.com*\n",
    "\n",
    "*Disclaimer 2: In case you wonder why I make use of horrible excel datasheets for this project: I blame path dependencies =D*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [],
   "source": [
    "# adjust your working directory accordingly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "D:\\Dropbox\\kasus\\workspace\\R\\pegida\\python\n"
     ]
    }
   ],
   "source": [
    "%cd \"D:\\Dropbox\\kasus\\workspace\\R\\pegida\\\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "import os\n",
    "import numpy as np\n",
    "from  statsmodels.tsa.stattools import grangercausalitytests\n",
    "import statsmodels.api as sm\n",
    "\n",
    "pal = sns.color_palette(\"Set1\", 9)[1:]\n",
    "palred = pal = sns.color_palette(\"Set1\", 9)\n",
    "sns.set(palette=pal, style='whitegrid', rc={'xtick.major.size': 5,\n",
    "                          'xtick.direction': 'inout'}, font_scale=1.1, font='sans-serif')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_proxy(p, s):\n",
    "    line = matplotlib.lines.Line2D([0], [0], linestyle=s, color=p)\n",
    "    return line\n",
    "\n",
    "def convert_date(row):\n",
    "    year = row.split('-')[0]\n",
    "    month = row.split('-')[1]\n",
    "    if len(month) == 1:\n",
    "        month = '0' + month\n",
    "    return year + '/' + month\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Main Paper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 520 entries, 2014-12-29 to 2016-05-31\n",
      "Data columns (total 4 columns):\n",
      " #   Column    Non-Null Count  Dtype\n",
      "---  ------    --------------  -----\n",
      " 0   posts     520 non-null    int64\n",
      " 1   likes     520 non-null    int64\n",
      " 2   comments  520 non-null    int64\n",
      " 3   shares    520 non-null    int64\n",
      "dtypes: int64(4)\n",
      "memory usage: 20.3+ KB\n"
     ]
    }
   ],
   "source": [
    "stats = pd.read_csv('data/pegida_fbstats_Dec14-May16.tsv', sep = \"\\t\", encoding='utf-8', index_col='day')\n",
    "stats.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-31</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-01</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares\n",
       "day                                       \n",
       "2014-12-29      1   4208      2547     885\n",
       "2014-12-30      1   8235     21882    2970\n",
       "2014-12-31      0      0         0       0\n",
       "2015-01-01      0      0         0       0\n",
       "2015-01-02      0      0         0       0"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate rolling averages\n",
    "stats['Posts'] = stats.posts.rolling(15, center=True).mean()\n",
    "stats['Comments'] = stats.comments.rolling(15, center=True).mean()\n",
    "stats['Likes'] = stats.likes.rolling(15, center=True).mean()\n",
    "stats2 = stats[['Posts', 'Comments']].dropna()\n",
    "stats3 = stats[['Posts', 'Likes']].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create dates for visualizations\n",
    "dates = ['2015-01-05',\n",
    "'2015-02-01',\n",
    "'2015-03-01',\n",
    "'2015-04-01',\n",
    "'2015-05-01',\n",
    "'2015-06-01',\n",
    "'2015-07-01',\n",
    "'2015-08-01',\n",
    "'2015-09-01',\n",
    "'2015-10-01',\n",
    "'2015-11-01',\n",
    "'2015-12-01',\n",
    "'2016-01-01',\n",
    "'2016-02-01',\n",
    "'2016-03-01',\n",
    "'2016-04-01',\n",
    "'2016-05-01']\n",
    "#'2016-05-25']\n",
    "locs = []\n",
    "for i in dates:\n",
    "    locs.append(stats2.index.get_loc(i))\n",
    "dates = [i[2:7].replace('-', '/')  for i in dates]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "charlie = stats2.index.get_loc('2015-01-07')\n",
    "cologne = stats2.index.get_loc('2016-01-01')\n",
    "jubilaeum = stats2.index.get_loc('2015-10-19')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlation: Posts and Comments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.09'"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original = str(stats.comments.corr(stats.posts))[:4]\n",
    "r_original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'-0.10'"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed =   str(stats2.Comments.corr(stats2.Posts))[:5]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlation: Posts and Likes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Posts</th>\n",
       "      <th>Likes</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-05</th>\n",
       "      <td>3.133333</td>\n",
       "      <td>15432.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-06</th>\n",
       "      <td>3.466667</td>\n",
       "      <td>16741.466667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Posts         Likes\n",
       "day                               \n",
       "2015-01-05  3.133333  15432.800000\n",
       "2015-01-06  3.466667  16741.466667"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats3.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.409'"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original = str(stats.likes.corr(stats.posts))[:5]\n",
    "r_original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.329'"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed =   str(stats3.Likes.corr(stats3.Posts))[:5]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Posts and Comments Aggregated per month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Likes</th>\n",
       "      <th>date</th>\n",
       "      <th>yearmonth</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares  Posts  Comments  Likes  \\\n",
       "day                                                                  \n",
       "2014-12-29      1   4208      2547     885    NaN       NaN    NaN   \n",
       "2014-12-30      1   8235     21882    2970    NaN       NaN    NaN   \n",
       "\n",
       "                  date yearmonth  \n",
       "day                               \n",
       "2014-12-29  2014-12-29   2014/12  \n",
       "2014-12-30  2014-12-30   2014/12  "
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats['date'] = stats.index\n",
    "\n",
    "stats['yearmonth'] = stats.date.apply(convert_date)\n",
    "stats.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>comments</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yearmonth</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014/12</th>\n",
       "      <td>2</td>\n",
       "      <td>24429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/01</th>\n",
       "      <td>130</td>\n",
       "      <td>273263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/02</th>\n",
       "      <td>140</td>\n",
       "      <td>108842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/03</th>\n",
       "      <td>132</td>\n",
       "      <td>76946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/04</th>\n",
       "      <td>90</td>\n",
       "      <td>42372</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           posts  comments\n",
       "yearmonth                 \n",
       "2014/12        2     24429\n",
       "2015/01      130    273263\n",
       "2015/02      140    108842\n",
       "2015/03      132     76946\n",
       "2015/04       90     42372"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_month = stats.groupby('yearmonth')[['posts', 'comments']].sum()\n",
    "stats_month.comments =  stats_month.comments.astype(int)\n",
    "stats_month.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grangertests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "news = pd.read_excel('data/pegida_news_Dec14-May16.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23882"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(news)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "newscounts = news.groupby('date').size()\n",
    "newscounts = newscounts.to_frame('newscounts')\n",
    "newscounts.index = newscounts.index.astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>newscounts</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-01</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-02</th>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-03</th>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-04</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-05</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            newscounts\n",
       "date                  \n",
       "2014-12-01           8\n",
       "2014-12-02           7\n",
       "2014-12-03          28\n",
       "2014-12-04           8\n",
       "2014-12-05           9"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "newscounts.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Likes</th>\n",
       "      <th>date</th>\n",
       "      <th>yearmonth</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-31</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-01</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015/01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>2015/01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares  Posts  Comments  Likes  \\\n",
       "day                                                                  \n",
       "2014-12-29      1   4208      2547     885    NaN       NaN    NaN   \n",
       "2014-12-30      1   8235     21882    2970    NaN       NaN    NaN   \n",
       "2014-12-31      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-01      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-02      0      0         0       0    NaN       NaN    NaN   \n",
       "\n",
       "                  date yearmonth  \n",
       "day                               \n",
       "2014-12-29  2014-12-29   2014/12  \n",
       "2014-12-30  2014-12-30   2014/12  \n",
       "2014-12-31  2014-12-31   2014/12  \n",
       "2015-01-01  2015-01-01   2015/01  \n",
       "2015-01-02  2015-01-02   2015/01  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Likes</th>\n",
       "      <th>date</th>\n",
       "      <th>yearmonth</th>\n",
       "      <th>newscounts</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>126.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>99.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-31</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>94.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-01</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015/01</td>\n",
       "      <td>15.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>2015/01</td>\n",
       "      <td>135.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares  Posts  Comments  Likes  \\\n",
       "day                                                                  \n",
       "2014-12-29      1   4208      2547     885    NaN       NaN    NaN   \n",
       "2014-12-30      1   8235     21882    2970    NaN       NaN    NaN   \n",
       "2014-12-31      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-01      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-02      0      0         0       0    NaN       NaN    NaN   \n",
       "\n",
       "                  date yearmonth  newscounts  \n",
       "day                                           \n",
       "2014-12-29  2014-12-29   2014/12       126.0  \n",
       "2014-12-30  2014-12-30   2014/12        99.0  \n",
       "2014-12-31  2014-12-31   2014/12        94.0  \n",
       "2015-01-01  2015-01-01   2015/01        15.0  \n",
       "2015-01-02  2015-01-02   2015/01       135.0  "
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_comb = stats.merge(newscounts, left_index= True, right_index=True, how='left')\n",
    "stats_comb.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Likes</th>\n",
       "      <th>date</th>\n",
       "      <th>yearmonth</th>\n",
       "      <th>newscounts</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>126.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>99.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares  Posts  Comments  Likes  \\\n",
       "day                                                                  \n",
       "2014-12-29      1   4208      2547     885    NaN       NaN    NaN   \n",
       "2014-12-30      1   8235     21882    2970    NaN       NaN    NaN   \n",
       "\n",
       "                  date yearmonth  newscounts  \n",
       "day                                           \n",
       "2014-12-29  2014-12-29   2014/12       126.0  \n",
       "2014-12-30  2014-12-30   2014/12        99.0  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_comb.newscounts.fillna(0, inplace=True)\n",
    "stats_comb.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "stats_comb['News'] = stats_comb.newscounts.rolling(15, center=True).mean()\n",
    "stats_comb['Comments'] = stats_comb.comments.rolling(15, center=True).mean()\n",
    "stats_comb['Posts'] = stats_comb.posts.rolling(15, center=True).mean()\n",
    "stats_comb['Likes'] = stats_comb.likes.rolling(15, center=True).mean()\n",
    "stats_comb['Shares'] = stats_comb.shares.rolling(15, center=True).mean()\n",
    "stats_comb['engagement'] = stats_comb['comments'] + stats_comb['likes'] + stats_comb['shares']\n",
    "stats_comb['Engagement'] = stats_comb.engagement.rolling(15, center = True).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>News</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Likes</th>\n",
       "      <th>Shares</th>\n",
       "      <th>Engagement</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-05-20</th>\n",
       "      <td>24.600000</td>\n",
       "      <td>1263.733333</td>\n",
       "      <td>10.200000</td>\n",
       "      <td>3499.866667</td>\n",
       "      <td>1194.600000</td>\n",
       "      <td>5958.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-21</th>\n",
       "      <td>26.200000</td>\n",
       "      <td>1343.600000</td>\n",
       "      <td>10.533333</td>\n",
       "      <td>3553.866667</td>\n",
       "      <td>1199.133333</td>\n",
       "      <td>6096.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-22</th>\n",
       "      <td>26.133333</td>\n",
       "      <td>1380.600000</td>\n",
       "      <td>10.933333</td>\n",
       "      <td>3766.600000</td>\n",
       "      <td>1256.733333</td>\n",
       "      <td>6403.933333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-23</th>\n",
       "      <td>28.266667</td>\n",
       "      <td>1394.133333</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>4174.400000</td>\n",
       "      <td>1311.266667</td>\n",
       "      <td>6879.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-24</th>\n",
       "      <td>28.200000</td>\n",
       "      <td>1465.866667</td>\n",
       "      <td>11.533333</td>\n",
       "      <td>4435.133333</td>\n",
       "      <td>1434.133333</td>\n",
       "      <td>7335.133333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 News     Comments      Posts        Likes       Shares  \\\n",
       "day                                                                       \n",
       "2016-05-20  24.600000  1263.733333  10.200000  3499.866667  1194.600000   \n",
       "2016-05-21  26.200000  1343.600000  10.533333  3553.866667  1199.133333   \n",
       "2016-05-22  26.133333  1380.600000  10.933333  3766.600000  1256.733333   \n",
       "2016-05-23  28.266667  1394.133333  11.000000  4174.400000  1311.266667   \n",
       "2016-05-24  28.200000  1465.866667  11.533333  4435.133333  1434.133333   \n",
       "\n",
       "             Engagement  \n",
       "day                      \n",
       "2016-05-20  5958.200000  \n",
       "2016-05-21  6096.600000  \n",
       "2016-05-22  6403.933333  \n",
       "2016-05-23  6879.800000  \n",
       "2016-05-24  7335.133333  "
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "statsc2 = stats_comb[['News', 'Comments', 'Posts', 'Likes', 'Shares', 'Engagement']].dropna()\n",
    "statsc3 = stats_comb[['News', 'Likes', 'Posts']].dropna()\n",
    "statsc2.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "Posts2 = stats_comb.posts.rolling(30, center=True).mean().dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "grangerposts = grangercausalitytests(statsc2[['Comments', 'Posts']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "granger_c_posts = grangercausalitytests(statsc2[['Posts', 'Comments']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "granger_n_posts = grangercausalitytests(statsc2[['Posts', 'News']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "granger_p_news = grangercausalitytests(statsc2[['News', 'Posts']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "grangernews = grangercausalitytests(statsc2[['Comments', 'News']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "granger_c_news = grangercausalitytests(statsc2[['News', 'Comments']], maxlag=7,\n",
    "                     verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "resultsposts = []\n",
    "for i in grangerposts.keys():\n",
    "    vals = {'F value P->C': round(grangerposts[i][0]['params_ftest'][0], 3), \n",
    "            'P value P->C':  round(grangerposts[i][0]['params_ftest'][1], 3)\n",
    "            , 'Lag' : i}\n",
    "    resultsposts.append(vals)\n",
    "    resposdf = pd.DataFrame(resultsposts)\n",
    "    resposdf.index = resposdf.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_c_posts = []\n",
    "for i in granger_c_posts.keys():\n",
    "    vals = {'F value C->P': round(granger_c_posts[i][0]['params_ftest'][0], 3), \n",
    "            'P value C->P':  round(granger_c_posts[i][0]['params_ftest'][1], 3)\n",
    "            , 'Lag' : i}\n",
    "    results_c_posts.append(vals)\n",
    "    respos_c_df = pd.DataFrame(results_c_posts)\n",
    "    respos_c_df.index = respos_c_df.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_n_posts = []\n",
    "for i in granger_n_posts.keys():\n",
    "    vals = {'F value N->P': round(granger_n_posts[i][0]['params_ftest'][0], 3), \n",
    "            'P value N->P':  round(granger_n_posts[i][0]['params_ftest'][1], 3)\n",
    "            , 'Lag' : i}\n",
    "    results_n_posts.append(vals)\n",
    "    respos_n_df = pd.DataFrame(results_n_posts)\n",
    "    respos_n_df.index = respos_n_df.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "resultsnews = []\n",
    "for i in grangernews.keys():\n",
    "    vals = {'F value N->C': round(grangernews[i][0]['params_ftest'][0], 3), \n",
    "            'P value N->C':  round(grangernews[i][0]['params_ftest'][1], 3), 'Lag' : i}\n",
    "    resultsnews.append(vals)\n",
    "    resnewsdf = pd.DataFrame(resultsnews)\n",
    "    resnewsdf.index = resnewsdf.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_c_news = []\n",
    "for i in granger_c_news.keys():\n",
    "    vals = {'F value C->N': round(granger_c_news[i][0]['params_ftest'][0], 3), \n",
    "            'P value C->N':  round(granger_c_news[i][0]['params_ftest'][1], 3), 'Lag' : i}\n",
    "    results_c_news.append(vals)\n",
    "    res_c_newsdf = pd.DataFrame(results_c_news)\n",
    "    res_c_newsdf.index = res_c_newsdf.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_p_news = []\n",
    "for i in granger_p_news.keys():\n",
    "    vals = {'F value P->N': round(granger_p_news[i][0]['params_ftest'][0], 3), \n",
    "            'P value P->N':  round(granger_p_news[i][0]['params_ftest'][1], 3), 'Lag' : i}\n",
    "    results_p_news.append(vals)\n",
    "    res_p_newsdf = pd.DataFrame(results_p_news)\n",
    "    res_p_newsdf.index = res_p_newsdf.Lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>P value P-&gt;C</th>\n",
       "      <th>P value C-&gt;P</th>\n",
       "      <th>P value P-&gt;N</th>\n",
       "      <th>P value N-&gt;P</th>\n",
       "      <th>P value N-&gt;C</th>\n",
       "      <th>P value C-&gt;N</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.327</td>\n",
       "      <td>0.585</td>\n",
       "      <td>0.144</td>\n",
       "      <td>0.712</td>\n",
       "      <td>0.276</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.471</td>\n",
       "      <td>0.443</td>\n",
       "      <td>0.576</td>\n",
       "      <td>0.852</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.135</td>\n",
       "      <td>0.421</td>\n",
       "      <td>0.864</td>\n",
       "      <td>0.724</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.139</td>\n",
       "      <td>0.223</td>\n",
       "      <td>0.953</td>\n",
       "      <td>0.824</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.166</td>\n",
       "      <td>0.126</td>\n",
       "      <td>0.896</td>\n",
       "      <td>0.886</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.136</td>\n",
       "      <td>0.111</td>\n",
       "      <td>0.377</td>\n",
       "      <td>0.741</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.161</td>\n",
       "      <td>0.162</td>\n",
       "      <td>0.300</td>\n",
       "      <td>0.818</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     P value P->C  P value C->P  P value P->N  P value N->P  P value N->C  \\\n",
       "Lag                                                                         \n",
       "1           0.327         0.585         0.144         0.712         0.276   \n",
       "2           0.471         0.443         0.576         0.852         0.044   \n",
       "3           0.135         0.421         0.864         0.724         0.000   \n",
       "4           0.139         0.223         0.953         0.824         0.000   \n",
       "5           0.166         0.126         0.896         0.886         0.000   \n",
       "6           0.136         0.111         0.377         0.741         0.000   \n",
       "7           0.161         0.162         0.300         0.818         0.000   \n",
       "\n",
       "     P value C->N  \n",
       "Lag                \n",
       "1             0.0  \n",
       "2             0.0  \n",
       "3             0.0  \n",
       "4             0.0  \n",
       "5             0.0  \n",
       "6             0.0  \n",
       "7             0.0  "
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grangerdf = pd.concat([resposdf, respos_c_df,  res_p_newsdf,  respos_n_df, resnewsdf,  res_c_newsdf,], axis=1)\n",
    "#grangerdf.index=grangerdf.Lag\n",
    "del grangerdf['Lag']\n",
    "tokeep = [i for i in grangerdf.columns if i.startswith('P')]\n",
    "grangerdf = grangerdf[tokeep]\n",
    "grangerdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrr}\n",
      "\\toprule\n",
      "{} &  P value P->C &  P value C->P &  P value P->N &  P value N->P &  P value N->C &  P value C->N \\\\\n",
      "Lag &               &               &               &               &               &               \\\\\n",
      "\\midrule\n",
      "1   &         0.327 &         0.585 &         0.144 &         0.712 &         0.276 &           0.0 \\\\\n",
      "2   &         0.471 &         0.443 &         0.576 &         0.852 &         0.044 &           0.0 \\\\\n",
      "3   &         0.135 &         0.421 &         0.864 &         0.724 &         0.000 &           0.0 \\\\\n",
      "4   &         0.139 &         0.223 &         0.953 &         0.824 &         0.000 &           0.0 \\\\\n",
      "5   &         0.166 &         0.126 &         0.896 &         0.886 &         0.000 &           0.0 \\\\\n",
      "6   &         0.136 &         0.111 &         0.377 &         0.741 &         0.000 &           0.0 \\\\\n",
      "7   &         0.161 &         0.162 &         0.300 &         0.818 &         0.000 &           0.0 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(grangerdf.to_latex())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Other correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.59'"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original = str(stats_comb.comments.corr(stats_comb.newscounts))[:4]\n",
    "r_original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.88'"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed = str(statsc2.Comments.corr(statsc2.News))[:4]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.46'"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original_all = str(stats_comb.engagement.corr(stats_comb.newscounts))[:4]\n",
    "r_original_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.28'"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(stats_comb.Engagement.corr(stats_comb.Posts))[:4]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.67'"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(stats_comb.Engagement.corr(stats_comb.News))[:4]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Likes</th>\n",
       "      <th>date</th>\n",
       "      <th>yearmonth</th>\n",
       "      <th>newscounts</th>\n",
       "      <th>News</th>\n",
       "      <th>Shares</th>\n",
       "      <th>engagement</th>\n",
       "      <th>Engagement</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-12-29</th>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>126.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7640</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-30</th>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>99.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>33087</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-12-31</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>2014/12</td>\n",
       "      <td>94.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-01</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015/01</td>\n",
       "      <td>15.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>2015/01</td>\n",
       "      <td>135.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-27</th>\n",
       "      <td>12</td>\n",
       "      <td>4343</td>\n",
       "      <td>1288</td>\n",
       "      <td>1408</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-05-27</td>\n",
       "      <td>2016/05</td>\n",
       "      <td>67.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7039</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-28</th>\n",
       "      <td>15</td>\n",
       "      <td>4408</td>\n",
       "      <td>2451</td>\n",
       "      <td>1590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-05-28</td>\n",
       "      <td>2016/05</td>\n",
       "      <td>36.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8449</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-29</th>\n",
       "      <td>12</td>\n",
       "      <td>4506</td>\n",
       "      <td>1178</td>\n",
       "      <td>1421</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-05-29</td>\n",
       "      <td>2016/05</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7105</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-30</th>\n",
       "      <td>8</td>\n",
       "      <td>8230</td>\n",
       "      <td>945</td>\n",
       "      <td>1671</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-05-30</td>\n",
       "      <td>2016/05</td>\n",
       "      <td>35.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10846</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-31</th>\n",
       "      <td>12</td>\n",
       "      <td>4800</td>\n",
       "      <td>1358</td>\n",
       "      <td>2215</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016-05-31</td>\n",
       "      <td>2016/05</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8373</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>520 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            posts  likes  comments  shares  Posts  Comments  Likes  \\\n",
       "day                                                                  \n",
       "2014-12-29      1   4208      2547     885    NaN       NaN    NaN   \n",
       "2014-12-30      1   8235     21882    2970    NaN       NaN    NaN   \n",
       "2014-12-31      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-01      0      0         0       0    NaN       NaN    NaN   \n",
       "2015-01-02      0      0         0       0    NaN       NaN    NaN   \n",
       "...           ...    ...       ...     ...    ...       ...    ...   \n",
       "2016-05-27     12   4343      1288    1408    NaN       NaN    NaN   \n",
       "2016-05-28     15   4408      2451    1590    NaN       NaN    NaN   \n",
       "2016-05-29     12   4506      1178    1421    NaN       NaN    NaN   \n",
       "2016-05-30      8   8230       945    1671    NaN       NaN    NaN   \n",
       "2016-05-31     12   4800      1358    2215    NaN       NaN    NaN   \n",
       "\n",
       "                  date yearmonth  newscounts  News  Shares  engagement  \\\n",
       "day                                                                      \n",
       "2014-12-29  2014-12-29   2014/12       126.0   NaN     NaN        7640   \n",
       "2014-12-30  2014-12-30   2014/12        99.0   NaN     NaN       33087   \n",
       "2014-12-31  2014-12-31   2014/12        94.0   NaN     NaN           0   \n",
       "2015-01-01  2015-01-01   2015/01        15.0   NaN     NaN           0   \n",
       "2015-01-02  2015-01-02   2015/01       135.0   NaN     NaN           0   \n",
       "...                ...       ...         ...   ...     ...         ...   \n",
       "2016-05-27  2016-05-27   2016/05        67.0   NaN     NaN        7039   \n",
       "2016-05-28  2016-05-28   2016/05        36.0   NaN     NaN        8449   \n",
       "2016-05-29  2016-05-29   2016/05        16.0   NaN     NaN        7105   \n",
       "2016-05-30  2016-05-30   2016/05        35.0   NaN     NaN       10846   \n",
       "2016-05-31  2016-05-31   2016/05         0.0   NaN     NaN        8373   \n",
       "\n",
       "            Engagement  \n",
       "day                     \n",
       "2014-12-29         NaN  \n",
       "2014-12-30         NaN  \n",
       "2014-12-31         NaN  \n",
       "2015-01-01         NaN  \n",
       "2015-01-02         NaN  \n",
       "...                ...  \n",
       "2016-05-27         NaN  \n",
       "2016-05-28         NaN  \n",
       "2016-05-29         NaN  \n",
       "2016-05-30         NaN  \n",
       "2016-05-31         NaN  \n",
       "\n",
       "[520 rows x 14 columns]"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_comb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "statsc2['Posts_rel'] = (statsc2['Posts'] - min(statsc2['Posts'])) / (max(statsc2['Posts']) - min(statsc2['Posts']))\n",
    "statsc2['News_rel'] = (statsc2['News'] - min(statsc2['News'])) / (max(statsc2['News']) - min(statsc2['News']))\n",
    "statsc2['Comments_rel'] = (statsc2['Comments'] - min(statsc2['Comments'])) / (max(statsc2['Comments']) - min(statsc2['Comments']))\n",
    "statsc2['Likes_rel'] = (statsc2['Likes'] - min(statsc2['Likes'])) / (max(statsc2['Likes']) - min(statsc2['Likes']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "statsc2['All'] = statsc2['Posts']  + statsc2['Likes'] + statsc2['Shares']\n",
    "statsc2['All_rel'] = (statsc2['All'] - min(statsc2['All'])) / (max(statsc2['All']) - min(statsc2['All']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.88'"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed = str(statsc2.Comments.corr(statsc2.News))[:4]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.56'"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " str(statsc2.Likes.corr(statsc2.News))[:4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.37'"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " str(statsc2.Shares.corr(statsc2.News))[:4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'-0.1'"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " str(statsc2.Posts.corr(statsc2.News))[:4]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Figure 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = statsc2[['Posts_rel' , 'Comments_rel', 'News_rel']].plot(rot=90, legend=False,\n",
    "                grid=False, color = [('#969696'), pal[0], pal[1]], figsize=(7,5), \n",
    "                                                              fontsize=12,\n",
    "                                                             ylim=(0,1),\n",
    "                                                             style =  ['-', '--', ':'])\n",
    "plt.xticks(locs ,dates)\n",
    "plt.tight_layout(pad=2)\n",
    "ax.set_ylabel('Normalized Activity', fontsize=14)\n",
    "ax.set_xlabel('Date', fontsize=14)\n",
    "sns.despine(right=False)\n",
    "\n",
    "ax.annotate('Hebdo\\nShooting', (charlie+25, 0.92 ), xytext=(25,-10), \n",
    "            textcoords='offset points', \n",
    "            arrowprops=dict(arrowstyle='-|>',\n",
    "                           facecolor='black'),\n",
    "           size=10)\n",
    "\n",
    "\n",
    "ax.annotate('Pegida\\nanniversary', (jubilaeum + 5, 0.35 ), xytext=(17,10), \n",
    "            textcoords='offset points', \n",
    "            arrowprops=dict(arrowstyle='-|>',\n",
    "                           facecolor='black'),\n",
    "            horizontalalignment='left',\n",
    "           size=10)\n",
    "\n",
    "ax.annotate('Cologne\\nAssaults', (cologne+15, 0.37 ), xytext=(25,0), \n",
    "            textcoords='offset points', \n",
    "            arrowprops=dict(arrowstyle='-|>',\n",
    "                           facecolor='black'),\n",
    "           size=10)\n",
    "\n",
    "\n",
    "plt.legend(labels =  ['Posts', 'Comments', 'News'], loc = 9,\n",
    "          fontsize = 11)\n",
    "ax.grid(True, alpha=0.3)\n",
    "\n",
    "ax1 = ax\n",
    "\n",
    "ax1.set_yticks(np.linspace(ax1.get_yticks()[0],ax1.get_yticks()[-1],len(ax1.get_yticks())))\n",
    "\n",
    "plt.savefig('pegida_fig1.jpeg', format='jpeg', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Appendix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### News and shares"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>News</th>\n",
       "      <th>Comments</th>\n",
       "      <th>Posts</th>\n",
       "      <th>Likes</th>\n",
       "      <th>Shares</th>\n",
       "      <th>Engagement</th>\n",
       "      <th>Posts_rel</th>\n",
       "      <th>News_rel</th>\n",
       "      <th>Comments_rel</th>\n",
       "      <th>Likes_rel</th>\n",
       "      <th>All</th>\n",
       "      <th>All_rel</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-05</th>\n",
       "      <td>140.466667</td>\n",
       "      <td>7835.266667</td>\n",
       "      <td>3.133333</td>\n",
       "      <td>15432.800000</td>\n",
       "      <td>2682.400000</td>\n",
       "      <td>25950.466667</td>\n",
       "      <td>0.086667</td>\n",
       "      <td>0.711584</td>\n",
       "      <td>0.734662</td>\n",
       "      <td>0.767638</td>\n",
       "      <td>18118.333333</td>\n",
       "      <td>0.774266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-06</th>\n",
       "      <td>150.133333</td>\n",
       "      <td>8317.333333</td>\n",
       "      <td>3.466667</td>\n",
       "      <td>16741.466667</td>\n",
       "      <td>2977.000000</td>\n",
       "      <td>28035.800000</td>\n",
       "      <td>0.120000</td>\n",
       "      <td>0.760554</td>\n",
       "      <td>0.782668</td>\n",
       "      <td>0.842658</td>\n",
       "      <td>19721.933333</td>\n",
       "      <td>0.852295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-07</th>\n",
       "      <td>157.666667</td>\n",
       "      <td>7604.533333</td>\n",
       "      <td>3.733333</td>\n",
       "      <td>16845.200000</td>\n",
       "      <td>3162.933333</td>\n",
       "      <td>27612.666667</td>\n",
       "      <td>0.146667</td>\n",
       "      <td>0.798717</td>\n",
       "      <td>0.711685</td>\n",
       "      <td>0.848604</td>\n",
       "      <td>20011.866667</td>\n",
       "      <td>0.866403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08</th>\n",
       "      <td>164.400000</td>\n",
       "      <td>8257.733333</td>\n",
       "      <td>4.133333</td>\n",
       "      <td>18285.866667</td>\n",
       "      <td>3520.400000</td>\n",
       "      <td>30064.000000</td>\n",
       "      <td>0.186667</td>\n",
       "      <td>0.832827</td>\n",
       "      <td>0.776733</td>\n",
       "      <td>0.931191</td>\n",
       "      <td>21810.400000</td>\n",
       "      <td>0.953917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-09</th>\n",
       "      <td>173.266667</td>\n",
       "      <td>8447.266667</td>\n",
       "      <td>4.200000</td>\n",
       "      <td>18388.866667</td>\n",
       "      <td>3560.933333</td>\n",
       "      <td>30397.066667</td>\n",
       "      <td>0.193333</td>\n",
       "      <td>0.877744</td>\n",
       "      <td>0.795608</td>\n",
       "      <td>0.937095</td>\n",
       "      <td>21954.000000</td>\n",
       "      <td>0.960904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-20</th>\n",
       "      <td>24.600000</td>\n",
       "      <td>1263.733333</td>\n",
       "      <td>10.200000</td>\n",
       "      <td>3499.866667</td>\n",
       "      <td>1194.600000</td>\n",
       "      <td>5958.200000</td>\n",
       "      <td>0.793333</td>\n",
       "      <td>0.124620</td>\n",
       "      <td>0.080245</td>\n",
       "      <td>0.083580</td>\n",
       "      <td>4704.666667</td>\n",
       "      <td>0.121575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-21</th>\n",
       "      <td>26.200000</td>\n",
       "      <td>1343.600000</td>\n",
       "      <td>10.533333</td>\n",
       "      <td>3553.866667</td>\n",
       "      <td>1199.133333</td>\n",
       "      <td>6096.600000</td>\n",
       "      <td>0.826667</td>\n",
       "      <td>0.132725</td>\n",
       "      <td>0.088198</td>\n",
       "      <td>0.086676</td>\n",
       "      <td>4763.533333</td>\n",
       "      <td>0.124440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-22</th>\n",
       "      <td>26.133333</td>\n",
       "      <td>1380.600000</td>\n",
       "      <td>10.933333</td>\n",
       "      <td>3766.600000</td>\n",
       "      <td>1256.733333</td>\n",
       "      <td>6403.933333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.132388</td>\n",
       "      <td>0.091883</td>\n",
       "      <td>0.098871</td>\n",
       "      <td>5034.266667</td>\n",
       "      <td>0.137613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-23</th>\n",
       "      <td>28.266667</td>\n",
       "      <td>1394.133333</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>4174.400000</td>\n",
       "      <td>1311.266667</td>\n",
       "      <td>6879.800000</td>\n",
       "      <td>0.873333</td>\n",
       "      <td>0.143195</td>\n",
       "      <td>0.093230</td>\n",
       "      <td>0.122248</td>\n",
       "      <td>5496.666667</td>\n",
       "      <td>0.160113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-05-24</th>\n",
       "      <td>28.200000</td>\n",
       "      <td>1465.866667</td>\n",
       "      <td>11.533333</td>\n",
       "      <td>4435.133333</td>\n",
       "      <td>1434.133333</td>\n",
       "      <td>7335.133333</td>\n",
       "      <td>0.926667</td>\n",
       "      <td>0.142857</td>\n",
       "      <td>0.100374</td>\n",
       "      <td>0.137195</td>\n",
       "      <td>5880.800000</td>\n",
       "      <td>0.178804</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>506 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  News     Comments      Posts         Likes       Shares  \\\n",
       "day                                                                         \n",
       "2015-01-05  140.466667  7835.266667   3.133333  15432.800000  2682.400000   \n",
       "2015-01-06  150.133333  8317.333333   3.466667  16741.466667  2977.000000   \n",
       "2015-01-07  157.666667  7604.533333   3.733333  16845.200000  3162.933333   \n",
       "2015-01-08  164.400000  8257.733333   4.133333  18285.866667  3520.400000   \n",
       "2015-01-09  173.266667  8447.266667   4.200000  18388.866667  3560.933333   \n",
       "...                ...          ...        ...           ...          ...   \n",
       "2016-05-20   24.600000  1263.733333  10.200000   3499.866667  1194.600000   \n",
       "2016-05-21   26.200000  1343.600000  10.533333   3553.866667  1199.133333   \n",
       "2016-05-22   26.133333  1380.600000  10.933333   3766.600000  1256.733333   \n",
       "2016-05-23   28.266667  1394.133333  11.000000   4174.400000  1311.266667   \n",
       "2016-05-24   28.200000  1465.866667  11.533333   4435.133333  1434.133333   \n",
       "\n",
       "              Engagement  Posts_rel  News_rel  Comments_rel  Likes_rel  \\\n",
       "day                                                                      \n",
       "2015-01-05  25950.466667   0.086667  0.711584      0.734662   0.767638   \n",
       "2015-01-06  28035.800000   0.120000  0.760554      0.782668   0.842658   \n",
       "2015-01-07  27612.666667   0.146667  0.798717      0.711685   0.848604   \n",
       "2015-01-08  30064.000000   0.186667  0.832827      0.776733   0.931191   \n",
       "2015-01-09  30397.066667   0.193333  0.877744      0.795608   0.937095   \n",
       "...                  ...        ...       ...           ...        ...   \n",
       "2016-05-20   5958.200000   0.793333  0.124620      0.080245   0.083580   \n",
       "2016-05-21   6096.600000   0.826667  0.132725      0.088198   0.086676   \n",
       "2016-05-22   6403.933333   0.866667  0.132388      0.091883   0.098871   \n",
       "2016-05-23   6879.800000   0.873333  0.143195      0.093230   0.122248   \n",
       "2016-05-24   7335.133333   0.926667  0.142857      0.100374   0.137195   \n",
       "\n",
       "                     All   All_rel  \n",
       "day                                 \n",
       "2015-01-05  18118.333333  0.774266  \n",
       "2015-01-06  19721.933333  0.852295  \n",
       "2015-01-07  20011.866667  0.866403  \n",
       "2015-01-08  21810.400000  0.953917  \n",
       "2015-01-09  21954.000000  0.960904  \n",
       "...                  ...       ...  \n",
       "2016-05-20   4704.666667  0.121575  \n",
       "2016-05-21   4763.533333  0.124440  \n",
       "2016-05-22   5034.266667  0.137613  \n",
       "2016-05-23   5496.666667  0.160113  \n",
       "2016-05-24   5880.800000  0.178804  \n",
       "\n",
       "[506 rows x 12 columns]"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "statsc2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.23'"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original = str(stats_comb.shares.corr(stats_comb.newscounts))[:4]\n",
    "r_original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.37'"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed = str(statsc2.Shares.corr(statsc2.News))[:4]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.38'"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_original = str(stats_comb.likes.corr(stats_comb.newscounts))[:4]\n",
    "r_original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.56'"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_smoothed = str(statsc3.Likes.corr(statsc3.News))[:4]\n",
    "r_smoothed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Media"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>date</th>\n",
       "      <th>length</th>\n",
       "      <th>title</th>\n",
       "      <th>text</th>\n",
       "      <th>source</th>\n",
       "      <th>code</th>\n",
       "      <th>pubtype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>393</td>\n",
       "      <td>\"Unheimliche Symbolik\"</td>\n",
       "      <td>Der Augenzeuge       Frank Richter,    54, ist...</td>\n",
       "      <td>Der Spiegel</td>\n",
       "      <td>SP</td>\n",
       "      <td>-99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>73</td>\n",
       "      <td>Aküfi</td>\n",
       "      <td>Als Nachwehen der Hogesa (Hooligans gegen Sala...</td>\n",
       "      <td>Frankfurter Rundschau</td>\n",
       "      <td>FR</td>\n",
       "      <td>-99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>620</td>\n",
       "      <td>Rechte Retter des Abendlandes</td>\n",
       "      <td>Ein Hogesa-Aktivist ruft zur Demo gegen Islami...</td>\n",
       "      <td>Frankfurter Rundschau</td>\n",
       "      <td>FR</td>\n",
       "      <td>-99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>302</td>\n",
       "      <td>Staus im Raum Dresden möglich</td>\n",
       "      <td>Heute werden sich erneut mehrere Tausend Mensc...</td>\n",
       "      <td>Sächsische Zeitung Regionalausgaben</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>458</td>\n",
       "      <td>Tillich fordert mehr Hilfe und Toleranz für Fl...</td>\n",
       "      <td>Sachsens Ministerpräsident erinnert daran, das...</td>\n",
       "      <td>Sächsische Zeitung Stammausgabe Dresden</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0       date  length  \\\n",
       "0           0 2014-12-01     393   \n",
       "1           1 2014-12-01      73   \n",
       "2           2 2014-12-01     620   \n",
       "3           3 2014-12-01     302   \n",
       "4           4 2014-12-01     458   \n",
       "\n",
       "                                               title  \\\n",
       "0                             \"Unheimliche Symbolik\"   \n",
       "1                                              Aküfi   \n",
       "2                      Rechte Retter des Abendlandes   \n",
       "3                      Staus im Raum Dresden möglich   \n",
       "4  Tillich fordert mehr Hilfe und Toleranz für Fl...   \n",
       "\n",
       "                                                text  \\\n",
       "0  Der Augenzeuge       Frank Richter,    54, ist...   \n",
       "1  Als Nachwehen der Hogesa (Hooligans gegen Sala...   \n",
       "2  Ein Hogesa-Aktivist ruft zur Demo gegen Islami...   \n",
       "3  Heute werden sich erneut mehrere Tausend Mensc...   \n",
       "4  Sachsens Ministerpräsident erinnert daran, das...   \n",
       "\n",
       "                                    source code  pubtype  \n",
       "0                              Der Spiegel   SP      -99  \n",
       "1                    Frankfurter Rundschau   FR      -99  \n",
       "2                    Frankfurter Rundschau   FR      -99  \n",
       "3      Sächsische Zeitung Regionalausgaben  -99  Zeitung  \n",
       "4  Sächsische Zeitung Stammausgabe Dresden  -99  Zeitung  "
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_date(row):\n",
    "    year = row.split('-')[0]\n",
    "    month = row.split('-')[1]\n",
    "    if len(month) == 1:\n",
    "        month = '0' + month\n",
    "    return year + '/' + month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "news['yearmonth'] = news.date.astype(str).apply(convert_date)\n",
    "news = news.sort_values('yearmonth')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>date</th>\n",
       "      <th>length</th>\n",
       "      <th>title</th>\n",
       "      <th>text</th>\n",
       "      <th>source</th>\n",
       "      <th>code</th>\n",
       "      <th>pubtype</th>\n",
       "      <th>yearmonth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>393</td>\n",
       "      <td>\"Unheimliche Symbolik\"</td>\n",
       "      <td>Der Augenzeuge       Frank Richter,    54, ist...</td>\n",
       "      <td>Der Spiegel</td>\n",
       "      <td>SP</td>\n",
       "      <td>-99</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1230</th>\n",
       "      <td>1230</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>1016</td>\n",
       "      <td>AUSGABE: HA</td>\n",
       "      <td>Herr Bedford-Strohm, haben die bayerischen Pro...</td>\n",
       "      <td>Nürnberger Nachrichten</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1229</th>\n",
       "      <td>1229</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>607</td>\n",
       "      <td>AUSGABE: HA</td>\n",
       "      <td>Wundern kann man sich schon, dass Pegida, dies...</td>\n",
       "      <td>Nürnberger Nachrichten</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1228</th>\n",
       "      <td>1228</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>860</td>\n",
       "      <td>LESERBRIEFE</td>\n",
       "      <td>Rechnung für das ÖlZum Bericht ,Ölpreis-Verfal...</td>\n",
       "      <td>Passauer Neue Presse (Stadt und Landkreis Passau)</td>\n",
       "      <td>a</td>\n",
       "      <td>-99</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1227</th>\n",
       "      <td>1227</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>363</td>\n",
       "      <td>Vom unheiligen Zorn zurück zur Vernunft</td>\n",
       "      <td>Kommentar  über den Skandal um den Kaufhaus-Re...</td>\n",
       "      <td>Sächsische Zeitung Stammausgabe Dresden</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0       date  length                                    title  \\\n",
       "0              0 2014-12-01     393                   \"Unheimliche Symbolik\"   \n",
       "1230        1230 2014-12-23    1016                              AUSGABE: HA   \n",
       "1229        1229 2014-12-23     607                              AUSGABE: HA   \n",
       "1228        1228 2014-12-23     860                              LESERBRIEFE   \n",
       "1227        1227 2014-12-23     363  Vom unheiligen Zorn zurück zur Vernunft   \n",
       "\n",
       "                                                   text  \\\n",
       "0     Der Augenzeuge       Frank Richter,    54, ist...   \n",
       "1230  Herr Bedford-Strohm, haben die bayerischen Pro...   \n",
       "1229  Wundern kann man sich schon, dass Pegida, dies...   \n",
       "1228  Rechnung für das ÖlZum Bericht ,Ölpreis-Verfal...   \n",
       "1227  Kommentar  über den Skandal um den Kaufhaus-Re...   \n",
       "\n",
       "                                                 source code  pubtype  \\\n",
       "0                                           Der Spiegel   SP      -99   \n",
       "1230                             Nürnberger Nachrichten  -99  Zeitung   \n",
       "1229                             Nürnberger Nachrichten  -99  Zeitung   \n",
       "1228  Passauer Neue Presse (Stadt und Landkreis Passau)    a      -99   \n",
       "1227            Sächsische Zeitung Stammausgabe Dresden  -99  Zeitung   \n",
       "\n",
       "     yearmonth  \n",
       "0      2014/12  \n",
       "1230   2014/12  \n",
       "1229   2014/12  \n",
       "1228   2014/12  \n",
       "1227   2014/12  "
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_agg = news.groupby('yearmonth')[['date']].count()\n",
    "news_agg.rename(columns = {'date': 'newscount'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>newscount</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yearmonth</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014/12</th>\n",
       "      <td>1836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/01</th>\n",
       "      <td>5698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/02</th>\n",
       "      <td>1934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/03</th>\n",
       "      <td>696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/04</th>\n",
       "      <td>1025</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           newscount\n",
       "yearmonth           \n",
       "2014/12         1836\n",
       "2015/01         5698\n",
       "2015/02         1934\n",
       "2015/03          696\n",
       "2015/04         1025"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_agg.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### FB and Media combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 520 entries, 0 to 519\n",
      "Data columns (total 5 columns):\n",
      " #   Column    Non-Null Count  Dtype \n",
      "---  ------    --------------  ----- \n",
      " 0   day       520 non-null    object\n",
      " 1   posts     520 non-null    int64 \n",
      " 2   likes     520 non-null    int64 \n",
      " 3   comments  520 non-null    int64 \n",
      " 4   shares    520 non-null    int64 \n",
      "dtypes: int64(4), object(1)\n",
      "memory usage: 20.4+ KB\n"
     ]
    }
   ],
   "source": [
    "stats = pd.read_csv('data/pegida_fbstats_Dec14-May16.tsv', sep = \"\\t\", encoding='utf-8')\n",
    "stats.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_date2(row):\n",
    "    year = row.split('-')[0]\n",
    "    month = row.split('-')[1]\n",
    "\n",
    "    return year + '/' + month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>day</th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>yearmonth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>1</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>885</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>1</td>\n",
       "      <td>8235</td>\n",
       "      <td>21882</td>\n",
       "      <td>2970</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          day  posts  likes  comments  shares yearmonth\n",
       "0  2014-12-29      1   4208      2547     885   2014/12\n",
       "1  2014-12-30      1   8235     21882    2970   2014/12"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats['yearmonth'] = stats.day.apply(convert_date2)\n",
    "stats.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yearmonth</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014/12</th>\n",
       "      <td>2</td>\n",
       "      <td>12443</td>\n",
       "      <td>24429</td>\n",
       "      <td>3855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/01</th>\n",
       "      <td>130</td>\n",
       "      <td>471913</td>\n",
       "      <td>273263</td>\n",
       "      <td>84014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/02</th>\n",
       "      <td>140</td>\n",
       "      <td>250578</td>\n",
       "      <td>108842</td>\n",
       "      <td>41498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/03</th>\n",
       "      <td>132</td>\n",
       "      <td>218534</td>\n",
       "      <td>76946</td>\n",
       "      <td>29373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/04</th>\n",
       "      <td>90</td>\n",
       "      <td>114038</td>\n",
       "      <td>42372</td>\n",
       "      <td>13337</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           posts   likes  comments  shares\n",
       "yearmonth                                 \n",
       "2014/12        2   12443     24429    3855\n",
       "2015/01      130  471913    273263   84014\n",
       "2015/02      140  250578    108842   41498\n",
       "2015/03      132  218534     76946   29373\n",
       "2015/04       90  114038     42372   13337"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats_agg = stats.groupby('yearmonth').sum()\n",
    "stats_agg.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = stats_agg.merge(news_agg, \n",
    "         right_index = True,left_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['posts'] = (df['posts'] / df['posts'].sum()) * 100\n",
    "df['likes'] = (df['likes'] / df['likes'].sum()) * 100\n",
    "df['comments'] = (df['comments'] / df['comments'].sum()) * 100\n",
    "\n",
    "df['newscount'] = (df['newscount'] / df['newscount'].sum()) * 100\n",
    "df['shares'] = (df['shares'] / df['shares'].max()) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>newscount</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yearmonth</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014/12</th>\n",
       "      <td>0.053121</td>\n",
       "      <td>0.301598</td>\n",
       "      <td>1.861403</td>\n",
       "      <td>3.462834</td>\n",
       "      <td>7.562091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/01</th>\n",
       "      <td>3.452855</td>\n",
       "      <td>11.438396</td>\n",
       "      <td>20.821672</td>\n",
       "      <td>75.467325</td>\n",
       "      <td>23.468841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/02</th>\n",
       "      <td>3.718459</td>\n",
       "      <td>6.073599</td>\n",
       "      <td>8.293375</td>\n",
       "      <td>37.276443</td>\n",
       "      <td>7.965732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/03</th>\n",
       "      <td>3.505976</td>\n",
       "      <td>5.296905</td>\n",
       "      <td>5.863012</td>\n",
       "      <td>26.384909</td>\n",
       "      <td>2.866675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015/04</th>\n",
       "      <td>2.390438</td>\n",
       "      <td>2.764094</td>\n",
       "      <td>3.228596</td>\n",
       "      <td>11.980238</td>\n",
       "      <td>4.221755</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              posts      likes   comments     shares  newscount\n",
       "yearmonth                                                      \n",
       "2014/12    0.053121   0.301598   1.861403   3.462834   7.562091\n",
       "2015/01    3.452855  11.438396  20.821672  75.467325  23.468841\n",
       "2015/02    3.718459   6.073599   8.293375  37.276443   7.965732\n",
       "2015/03    3.505976   5.296905   5.863012  26.384909   2.866675\n",
       "2015/04    2.390438   2.764094   3.228596  11.980238   4.221755"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Durchgezaehlt.org"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This code likely won't run in the future as it scrapes the data from the website durchgezaehlt.org"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "dfd = pd.read_html('https://durchgezaehlt.org/pegida-dresden-statistik/', flavor='html5lib', header=0)[0]\n",
    "dfd.rename(columns= {'Datum': 'date', 'Durchgezählt (Minimum)*': 'durchgezaehlt_min',\n",
    "                   'Durchgezählt (Maximum)*': 'durchgezaehlt_max', \n",
    "                    'Polizei': 'police', 'Presse **': 'press', \n",
    "                   'Prof. Rucht et al.': 'rucht_et_al'}, inplace=True)\n",
    "\n",
    "def cutter(row):\n",
    "    return row.split()[0]\n",
    "def get_monthyear3(row):\n",
    "    return row.split(\".\")[2]+ '/' + row.split(\".\")[1]\n",
    "\n",
    "dfd.date = dfd.date.astype(str).apply(cutter)\n",
    "dfd.durchgezaehlt_min = dfd.durchgezaehlt_min.astype(str).apply(cutter)\n",
    "dfd.durchgezaehlt_max = dfd.durchgezaehlt_max.astype(str).apply(cutter)\n",
    "dfd.police = dfd.police.astype(str).apply(cutter)\n",
    "dfd.press = dfd.press.astype(str).apply(cutter)\n",
    "dfd.rucht_et_al = dfd.rucht_et_al.astype(str).apply(cutter)\n",
    "\n",
    "dfd.durchgezaehlt_min = pd.to_numeric(dfd.durchgezaehlt_min, errors='coerce')\n",
    "dfd.durchgezaehlt_max = pd.to_numeric(dfd.durchgezaehlt_max, errors='coerce')\n",
    "dfd.police = pd.to_numeric(dfd.police, errors='coerce')\n",
    "dfd.press = pd.to_numeric(dfd.press, errors='coerce')\n",
    "dfd.rucht_et_al = pd.to_numeric(dfd.rucht_et_al, errors='coerce')\n",
    "\n",
    "dfd['nr_sources'] = dfd[['durchgezaehlt_min', 'durchgezaehlt_max', 'police', 'press',\n",
    "       'rucht_et_al']].count(axis=1)\n",
    "dfd['avg_nr'] = dfd[['durchgezaehlt_min', 'durchgezaehlt_max', 'police', 'press',\n",
    "       'rucht_et_al']].sum(axis=1) / dfd['nr_sources']\n",
    "dfd['date_day'] = pd.to_datetime(dfd['date'], format='%d.%m.%Y')\n",
    "dfd['date'] = dfd.date.apply(get_monthyear3)\n",
    "dfd.index = dfd.date\n",
    "del dfd['date']\n",
    "\n",
    "dfd.to_excel('data/pegida_democounts.xlsx', encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>durchgezaehlt_min</th>\n",
       "      <th>durchgezaehlt_max</th>\n",
       "      <th>police</th>\n",
       "      <th>press</th>\n",
       "      <th>rucht_et_al</th>\n",
       "      <th>nr_sources</th>\n",
       "      <th>avg_nr</th>\n",
       "      <th>date_day</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016/05</th>\n",
       "      <td>2016/05</td>\n",
       "      <td>2900.0</td>\n",
       "      <td>3500.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>3200.0</td>\n",
       "      <td>2016-05-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016/05</th>\n",
       "      <td>2016/05</td>\n",
       "      <td>2500.0</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2750.0</td>\n",
       "      <td>2016-05-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016/05</th>\n",
       "      <td>2016/05</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>2800.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2400.0</td>\n",
       "      <td>2016-05-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016/05</th>\n",
       "      <td>2016/05</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>2500.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2250.0</td>\n",
       "      <td>2016-05-23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016/05</th>\n",
       "      <td>2016/05</td>\n",
       "      <td>2500.0</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2750.0</td>\n",
       "      <td>2016-05-30</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  durchgezaehlt_min  durchgezaehlt_max  police  press  \\\n",
       "date                                                                    \n",
       "2016/05  2016/05             2900.0             3500.0     NaN    NaN   \n",
       "2016/05  2016/05             2500.0             3000.0     NaN    NaN   \n",
       "2016/05  2016/05             2000.0             2800.0     NaN    NaN   \n",
       "2016/05  2016/05             2000.0             2500.0     NaN    NaN   \n",
       "2016/05  2016/05             2500.0             3000.0     NaN    NaN   \n",
       "\n",
       "         rucht_et_al  nr_sources  avg_nr   date_day  \n",
       "date                                                 \n",
       "2016/05          NaN           2  3200.0 2016-05-02  \n",
       "2016/05          NaN           2  2750.0 2016-05-09  \n",
       "2016/05          NaN           2  2400.0 2016-05-16  \n",
       "2016/05          NaN           2  2250.0 2016-05-23  \n",
       "2016/05          NaN           2  2750.0 2016-05-30  "
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfd = pd.read_excel('data/pegida_democounts.xlsx')\n",
    "dfd = dfd[dfd.date_day < '2016-06-01']\n",
    "dfd.index = dfd.date\n",
    "dfd.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>avg_demo</th>\n",
       "      <th>posts</th>\n",
       "      <th>likes</th>\n",
       "      <th>comments</th>\n",
       "      <th>shares</th>\n",
       "      <th>newscount</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14/12</th>\n",
       "      <td>12500.000000</td>\n",
       "      <td>0.053121</td>\n",
       "      <td>0.301598</td>\n",
       "      <td>1.861403</td>\n",
       "      <td>3.462834</td>\n",
       "      <td>7.562091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15/01</th>\n",
       "      <td>18766.666667</td>\n",
       "      <td>3.452855</td>\n",
       "      <td>11.438396</td>\n",
       "      <td>20.821672</td>\n",
       "      <td>75.467325</td>\n",
       "      <td>23.468841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15/02</th>\n",
       "      <td>3700.000000</td>\n",
       "      <td>3.718459</td>\n",
       "      <td>6.073599</td>\n",
       "      <td>8.293375</td>\n",
       "      <td>37.276443</td>\n",
       "      <td>7.965732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15/03</th>\n",
       "      <td>5566.666667</td>\n",
       "      <td>3.505976</td>\n",
       "      <td>5.296905</td>\n",
       "      <td>5.863012</td>\n",
       "      <td>26.384909</td>\n",
       "      <td>2.866675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15/04</th>\n",
       "      <td>6394.444444</td>\n",
       "      <td>2.390438</td>\n",
       "      <td>2.764094</td>\n",
       "      <td>3.228596</td>\n",
       "      <td>11.980238</td>\n",
       "      <td>4.221755</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           avg_demo     posts      likes   comments     shares  newscount\n",
       "date                                                                     \n",
       "14/12  12500.000000  0.053121   0.301598   1.861403   3.462834   7.562091\n",
       "15/01  18766.666667  3.452855  11.438396  20.821672  75.467325  23.468841\n",
       "15/02   3700.000000  3.718459   6.073599   8.293375  37.276443   7.965732\n",
       "15/03   5566.666667  3.505976   5.296905   5.863012  26.384909   2.866675\n",
       "15/04   6394.444444  2.390438   2.764094   3.228596  11.980238   4.221755"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_agg = dfd.groupby(dfd.index).mean()[2:]\n",
    "count_agg.rename(columns={'avg_nr': 'avg_demo'}, inplace=True)\n",
    "\n",
    "count_agg = count_agg[['avg_demo']].merge(df, left_index=True, right_index=True)\n",
    "count_agg.index = count_agg.index.str[2:]\n",
    "count_agg.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_agg['avg_demo_std'] = count_agg['avg_demo'] / count_agg['avg_demo'].sum()\n",
    "count_agg['comments_std'] = count_agg['comments'] / count_agg['comments'].sum()\n",
    "count_agg['newscount_std'] = count_agg['newscount'] / count_agg['newscount'].sum()\n",
    "count_agg['posts_std'] = count_agg['posts'] / count_agg['posts'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_agg['news_rel'] = (count_agg['newscount'] - min(count_agg['newscount'])) / (max(count_agg['newscount']) - min(count_agg['newscount']))\n",
    "count_agg['comments_rel'] = (count_agg['comments'] - min(count_agg['comments'])) / (max(count_agg['comments']) - min(count_agg['comments']))\n",
    "count_agg['demos_rel'] = (count_agg['avg_demo'] - min(count_agg['avg_demo'])) / (max(count_agg['avg_demo']) - min(count_agg['avg_demo']))\n",
    "count_agg['likes_rel'] = (count_agg['likes'] - min(count_agg['likes'])) / (max(count_agg['likes']) - min(count_agg['likes']))\n",
    "count_agg['shares_rel'] = (count_agg['shares'] - min(count_agg['shares'])) / (max(count_agg['shares']) - min(count_agg['shares']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pal = sns.color_palette(\"Set1\", 6)[0:]\n",
    "sns.set(palette = pal, style = 'whitegrid', rc = {'xtick.major.size': 5,\n",
    "                          'xtick.direction': 'inout'}, font_scale = 1.1 , \n",
    "                          font = 'sans-serif')\n",
    "\n",
    "g = count_agg[[ 'comments_rel', 'likes_rel', 'shares_rel',   'news_rel', 'demos_rel',]].plot( rot=90,  \n",
    "                 fontsize = 10,   lw = 2,\n",
    "                legend = False,figsize = (9,6),\n",
    "                marker = '^', markersize = 8, ylim = (0,1))\n",
    "\n",
    "\n",
    "\n",
    "line1 = g.lines[0]\n",
    "line2 = g.lines[1]\n",
    "line3 = g.lines[2]\n",
    "line4 = g.lines[3]\n",
    "line5 = g.lines[4]\n",
    "line1.set_marker('o')\n",
    "line2.set_marker('s')\n",
    "line3.set_marker('D')\n",
    "line4.set_marker('v')\n",
    "line5.set_marker('X')\n",
    "\n",
    "g.set_xlabel('Date', fontsize=13)\n",
    "g.set_ylabel('Normalized Activity', fontsize = 13)\n",
    "plt.yticks(g.get_yticks(), ['{0:.2f}'.format(i )  for i in g.get_yticks()])\n",
    "\n",
    "plt.legend([line1, line2, line3, line4, line5], ['Comments', 'Likes', 'Shares', 'News', \n",
    "                                  'Demonstrations'])\n",
    "\n",
    "plt.savefig('pegida_suppfig_engagementnewsdemos.jpeg', format='jpeg', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### News Sources"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import nltk\n",
    "\n",
    "import seaborn as sns\n",
    "sns.set(style='whitegrid')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>date</th>\n",
       "      <th>length</th>\n",
       "      <th>title</th>\n",
       "      <th>text</th>\n",
       "      <th>source</th>\n",
       "      <th>code</th>\n",
       "      <th>pubtype</th>\n",
       "      <th>yearmonth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2014-12-01</td>\n",
       "      <td>393</td>\n",
       "      <td>\"Unheimliche Symbolik\"</td>\n",
       "      <td>Der Augenzeuge       Frank Richter,    54, ist Direktor der Sächsischen Landeszentrale für politische Bildung. Zufällig traf er in Dresden auf eine Demonstration der \"Patriotischen Europäer gegen die Islamisierung des Abendlandes\" (Pegida), die seit Wochen montags auf die Straße gehen. Zuletzt w...</td>\n",
       "      <td>Der Spiegel</td>\n",
       "      <td>SP</td>\n",
       "      <td>-99</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1230</th>\n",
       "      <td>1230</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>1016</td>\n",
       "      <td>AUSGABE: HA</td>\n",
       "      <td>Herr Bedford-Strohm, haben die bayerischen Protestanten seit Ihrer Wahl zum EKD-Ratsvorsitzenden nur noch einen halben Landesbischof?  Heinrich Bedford-Strohm: Es gibt einen ganzen Landesbischof und einen ganzen EKD-Ratsvorsitzenden. Der muss jetzt allerdings die Termine noch besser gegeneinande...</td>\n",
       "      <td>Nürnberger Nachrichten</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1229</th>\n",
       "      <td>1229</td>\n",
       "      <td>2014-12-23</td>\n",
       "      <td>607</td>\n",
       "      <td>AUSGABE: HA</td>\n",
       "      <td>Wundern kann man sich schon, dass Pegida, diese schwierige Mischung aus Rechtsextremen und stockkonservativem Wutbürgertum, das Abendland ausgerechnet in Sachsen vor der Islamisierung retten will. Die Zahl der Muslime dort erreicht allenfalls ein halbes Prozent; Überfremdung muss man sich wohl a...</td>\n",
       "      <td>Nürnberger Nachrichten</td>\n",
       "      <td>-99</td>\n",
       "      <td>Zeitung</td>\n",
       "      <td>2014/12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0       date  length                   title  \\\n",
       "0              0 2014-12-01     393  \"Unheimliche Symbolik\"   \n",
       "1230        1230 2014-12-23    1016             AUSGABE: HA   \n",
       "1229        1229 2014-12-23     607             AUSGABE: HA   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                             text  \\\n",
       "0     Der Augenzeuge       Frank Richter,    54, ist Direktor der Sächsischen Landeszentrale für politische Bildung. Zufällig traf er in Dresden auf eine Demonstration der \"Patriotischen Europäer gegen die Islamisierung des Abendlandes\" (Pegida), die seit Wochen montags auf die Straße gehen. Zuletzt w...   \n",
       "1230  Herr Bedford-Strohm, haben die bayerischen Protestanten seit Ihrer Wahl zum EKD-Ratsvorsitzenden nur noch einen halben Landesbischof?  Heinrich Bedford-Strohm: Es gibt einen ganzen Landesbischof und einen ganzen EKD-Ratsvorsitzenden. Der muss jetzt allerdings die Termine noch besser gegeneinande...   \n",
       "1229  Wundern kann man sich schon, dass Pegida, diese schwierige Mischung aus Rechtsextremen und stockkonservativem Wutbürgertum, das Abendland ausgerechnet in Sachsen vor der Islamisierung retten will. Die Zahl der Muslime dort erreicht allenfalls ein halbes Prozent; Überfremdung muss man sich wohl a...   \n",
       "\n",
       "                      source code  pubtype yearmonth  \n",
       "0                Der Spiegel   SP      -99   2014/12  \n",
       "1230  Nürnberger Nachrichten  -99  Zeitung   2014/12  \n",
       "1229  Nürnberger Nachrichten  -99  Zeitung   2014/12  "
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.max_colwidth', 300)\n",
    "news.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [],
   "source": [
    "sources = news.source.value_counts().reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "116"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(sources)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lr}\n",
      "\\toprule\n",
      "                                       index &  source \\\\\n",
      "\\midrule\n",
      "         Sächsische Zeitung Regionalausgaben &    2243 \\\\\n",
      "     Sächsische Zeitung Stammausgabe Dresden &    1330 \\\\\n",
      "                       Frankfurter Rundschau &    1069 \\\\\n",
      "                       taz, die tageszeitung &     967 \\\\\n",
      "                             dpa-AFX ProFeed &     905 \\\\\n",
      "              Agence France Presse -- German &     896 \\\\\n",
      "                 Rheinische Post Duesseldorf &     895 \\\\\n",
      "                            Berliner Zeitung &     773 \\\\\n",
      "                            Der Tagesspiegel &     717 \\\\\n",
      "  Frankfurter Neue Presse (Regionalausgaben) &     684 \\\\\n",
      " abendblatt.de - Hamburger Abendblatt Online &     622 \\\\\n",
      "                       Kölner Stadt-Anzeiger &     606 \\\\\n",
      "                      Nürnberger Nachrichten &     582 \\\\\n",
      "                      Mitteldeutsche Zeitung &     533 \\\\\n",
      "                         Stuttgarter Zeitung &     483 \\\\\n",
      "                       WELT ONLINE (Deutsch) &     455 \\\\\n",
      "                        Aachener Nachrichten &     434 \\\\\n",
      "                  Berliner Morgenpost Online &     434 \\\\\n",
      "                            Aachener Zeitung &     426 \\\\\n",
      "                     Stuttgarter Nachrichten &     425 \\\\\n",
      "                          Nürnberger Zeitung &     422 \\\\\n",
      "                         Kölnische Rundschau &     417 \\\\\n",
      "                     General-Anzeiger (Bonn) &     415 \\\\\n",
      "                                    Die Welt &     389 \\\\\n",
      "                              Südwest Presse &     374 \\\\\n",
      "                        Hamburger Abendblatt &     373 \\\\\n",
      "                              Kölner Express &     365 \\\\\n",
      "                         Berliner Morgenpost &     361 \\\\\n",
      "                   SDA - Basisdienst Deutsch &     324 \\\\\n",
      "                              SPIEGEL ONLINE &     293 \\\\\n",
      "                             Berliner Kurier &     263 \\\\\n",
      "                Allgemeine Zeitung (Germany) &     247 \\\\\n",
      "                                 ZEIT-online &     230 \\\\\n",
      "                                Welt kompakt &     222 \\\\\n",
      "                                        B.Z. &     198 \\\\\n",
      "              Wiesbadener Tagblatt (Germany) &     158 \\\\\n",
      "           Die ZEIT (inklusive ZEIT Magazin) &     158 \\\\\n",
      "                Wiesbadener Kurier (Germany) &     156 \\\\\n",
      "                Main-Taunus-Kurier (Germany) &     146 \\\\\n",
      "                          Aar-Bote (Germany) &     146 \\\\\n",
      "                 Idsteiner Zeitung (Germany) &     144 \\\\\n",
      "                       Main-Spitze (Germany) &     144 \\\\\n",
      "                   Wormser Zeitung (Germany) &     143 \\\\\n",
      "                Neuss Grevenbroicher Zeitung &     137 \\\\\n",
      "                Giessener Anzeiger (Germany) &     132 \\\\\n",
      "                        Bergische Morgenpost &     125 \\\\\n",
      "             Lampertheimer Zeitung (Germany) &     121 \\\\\n",
      "                Bürstädter Zeitung (Germany) &     121 \\\\\n",
      "                                 Der Spiegel &     121 \\\\\n",
      "                         Solinger Morgenpost &     119 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(sources.head(50).to_latex(index=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### News - Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import sklearn\n",
    "import random\n",
    "import nltk\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.svm import LinearSVC\n",
    "from datetime import datetime\n",
    "from nltk.corpus import stopwords\n",
    "import Stemmer # on windows: build wheel\n",
    "\n",
    "class StemmedTfidfVectorizer(TfidfVectorizer):\n",
    "     def build_analyzer(self):\n",
    "         analyzer = super(TfidfVectorizer, self).build_analyzer()\n",
    "         return lambda doc: german_stemmer.stemWords(analyzer(doc))\n",
    "\n",
    "random.seed(1337)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "246"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "swords = stopwords.words('german')\n",
    "german_stemmer = Stemmer.Stemmer('de')\n",
    "swords += ['januar', 'februar', 'märz', 'april', 'mai', 'juni',\n",
    "          'juli', 'august', 'september', 'oktober', 'november', 'dezember',\n",
    "          '2014', '2015', '2016']\n",
    "len(swords)       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "def printf(model, class_, nr_,vocabulary, rev=True):\n",
    "    important_features = list(zip(vocabulary, model.coef_[class_]))\n",
    "    sort_ = sorted(important_features, key = lambda x : x[1], reverse=rev)[:nr_]\n",
    "    return(sort_)\n",
    "\n",
    "\n",
    "def date_label(row, cutoffs):\n",
    "    cutoffs = [datetime.strptime(d, \"%Y-%m-%d\") for d in cutoffs]\n",
    "    labels = range(0, len(cutoffs))\n",
    "    for i in labels:\n",
    "        if row >= cutoffs[i] and row < cutoffs[i+1]:\n",
    "            return i\n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = news.text.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9790"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transformer = StemmedTfidfVectorizer(min_df = 50, stop_words = swords,\n",
    "                             ngram_range = (1,1),\n",
    "                             analyzer = 'word').fit(n)\n",
    "vocab = transformer.get_feature_names()\n",
    "dtm = transformer.transform(n)\n",
    "len(vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    9468\n",
       "4    4251\n",
       "3    4086\n",
       "5    2499\n",
       "1    2333\n",
       "2    1642\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cutoffs = ['2014-12-01',  \n",
    "           '2015-03-01',\n",
    "           '2015-06-01', \n",
    "           '2015-09-01',\n",
    "           '2015-12-01',\n",
    "           '2016-03-01',\n",
    "           '2016-05-31']\n",
    "\n",
    "news['label'] = news.date.apply(date_label, args=(cutoffs,))\n",
    "labels = news.label.values\n",
    "news.label.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = LinearSVC(fit_intercept = True, penalty = 'l2', multi_class = \"crammer_singer\" ,\n",
    "                class_weight = \"balanced\") # \n",
    "clf.fit(dtm, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_labels =  []\n",
    "for i in clf.classes_:\n",
    "    comb =printf(clf, i, 10, vocab, rev=True)\n",
    "    df = pd.DataFrame(comb, columns =[cutoffs[i] + ' / ' + cutoffs[i+1], 'importance'])\n",
    "    df['rank'] =  df.index + 1\n",
    "    df = df.set_index(['rank'])\n",
    "    del df['importance']\n",
    "    news_labels.append(df)\n",
    "df = pd.concat(news_labels, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>2014-12-01 / 2015-03-01</th>\n",
       "      <th>2015-03-01 / 2015-06-01</th>\n",
       "      <th>2015-06-01 / 2015-09-01</th>\n",
       "      <th>2015-09-01 / 2015-12-01</th>\n",
       "      <th>2015-12-01 / 2016-03-01</th>\n",
       "      <th>2016-03-01 / 2016-05-31</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rank</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>neujahrsempfang</td>\n",
       "      <td>troglitz</td>\n",
       "      <td>verfassungsschutzbericht</td>\n",
       "      <td>galg</td>\n",
       "      <td>silvesternacht</td>\n",
       "      <td>afd</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>pegida</td>\n",
       "      <td>geert</td>\n",
       "      <td>fluchtlingsheim</td>\n",
       "      <td>jahrestag</td>\n",
       "      <td>fluchtlingskris</td>\n",
       "      <td>katholikentag</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>islamkrit</td>\n",
       "      <td>wuppertal</td>\n",
       "      <td>heidenau</td>\n",
       "      <td>fluchtlingskris</td>\n",
       "      <td>silv</td>\n",
       "      <td>kinderschokolad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>charli</td>\n",
       "      <td>wild</td>\n",
       "      <td>fluchtlingsunterkunft</td>\n",
       "      <td>1938</td>\n",
       "      <td>russlanddeutsch</td>\n",
       "      <td>clausnitz</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>islamisier</td>\n",
       "      <td>islamkrit</td>\n",
       "      <td>freital</td>\n",
       "      <td>transitzon</td>\n",
       "      <td>clausnitz</td>\n",
       "      <td>jena</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>kathrin</td>\n",
       "      <td>gey</td>\n",
       "      <td>austritt</td>\n",
       "      <td>rek</td>\n",
       "      <td>connewitz</td>\n",
       "      <td>bohmermann</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>demonstration</td>\n",
       "      <td>befreiung</td>\n",
       "      <td>ramadan</td>\n",
       "      <td>einjahr</td>\n",
       "      <td>europaweit</td>\n",
       "      <td>fluchtlingskris</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>kogida</td>\n",
       "      <td>luck</td>\n",
       "      <td>alfa</td>\n",
       "      <td>gift</td>\n",
       "      <td>59</td>\n",
       "      <td>hof</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>abendland</td>\n",
       "      <td>blockupy</td>\n",
       "      <td>jag</td>\n",
       "      <td>asylchaos</td>\n",
       "      <td>obergrenz</td>\n",
       "      <td>geldstraf</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ukrain</td>\n",
       "      <td>henkel</td>\n",
       "      <td>zeltstadt</td>\n",
       "      <td>schaff</td>\n",
       "      <td>warschau</td>\n",
       "      <td>hattk</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     2014-12-01 / 2015-03-01 2015-03-01 / 2015-06-01  \\\n",
       "rank                                                   \n",
       "1            neujahrsempfang                troglitz   \n",
       "2                     pegida                   geert   \n",
       "3                  islamkrit               wuppertal   \n",
       "4                     charli                    wild   \n",
       "5                 islamisier               islamkrit   \n",
       "6                    kathrin                     gey   \n",
       "7              demonstration               befreiung   \n",
       "8                     kogida                    luck   \n",
       "9                  abendland                blockupy   \n",
       "10                    ukrain                  henkel   \n",
       "\n",
       "       2015-06-01 / 2015-09-01 2015-09-01 / 2015-12-01  \\\n",
       "rank                                                     \n",
       "1     verfassungsschutzbericht                    galg   \n",
       "2              fluchtlingsheim               jahrestag   \n",
       "3                     heidenau         fluchtlingskris   \n",
       "4        fluchtlingsunterkunft                    1938   \n",
       "5                      freital              transitzon   \n",
       "6                     austritt                     rek   \n",
       "7                      ramadan                 einjahr   \n",
       "8                         alfa                    gift   \n",
       "9                          jag               asylchaos   \n",
       "10                   zeltstadt                  schaff   \n",
       "\n",
       "     2015-12-01 / 2016-03-01 2016-03-01 / 2016-05-31  \n",
       "rank                                                  \n",
       "1             silvesternacht                     afd  \n",
       "2            fluchtlingskris           katholikentag  \n",
       "3                       silv         kinderschokolad  \n",
       "4            russlanddeutsch               clausnitz  \n",
       "5                  clausnitz                    jena  \n",
       "6                  connewitz              bohmermann  \n",
       "7                 europaweit         fluchtlingskris  \n",
       "8                         59                     hof  \n",
       "9                  obergrenz               geldstraf  \n",
       "10                  warschau                   hattk  "
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2014-12-01\n",
      "[('neujahrsempfang', 2.94614821870359), ('pegida', 2.554780952541824), ('islamkrit', 2.5483995491966978), ('charli', 2.068904650461388), ('islamisier', 2.0455301260525856)]\n",
      "2015-03-01\n",
      "[('troglitz', 4.032256231462412), ('geert', 2.9588447043545085), ('wuppertal', 2.309441315407757), ('wild', 1.9849163394592513), ('islamkrit', 1.7710009642242026)]\n",
      "2015-06-01\n",
      "[('verfassungsschutzbericht', 1.79684992877898), ('fluchtlingsheim', 1.660634531845579), ('heidenau', 1.6604194025032613), ('fluchtlingsunterkunft', 1.6517165370978992), ('freital', 1.6169033173281084)]\n",
      "2015-09-01\n",
      "[('galg', 4.876430037204953), ('jahrestag', 2.6740654878654193), ('fluchtlingskris', 2.670427929778026), ('1938', 2.2513904169981633), ('transitzon', 2.234375904834801)]\n",
      "2015-12-01\n",
      "[('silvesternacht', 3.415492416234084), ('fluchtlingskris', 2.691105610314481), ('silv', 2.5818460909752425), ('russlanddeutsch', 2.199628600451065), ('clausnitz', 2.0661008568915324)]\n",
      "2016-03-01\n",
      "[('afd', 2.3309761998722944), ('katholikentag', 2.1329446843011506), ('kinderschokolad', 1.968922819642008), ('clausnitz', 1.8550853646288166), ('jena', 1.7701579297564225)]\n"
     ]
    }
   ],
   "source": [
    "for i in clf.classes_:\n",
    "    print(cutoffs[i])\n",
    "    print(printf(clf, i, 5, vocab, rev=True))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Comment activity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from IPython.display import clear_output\n",
    "import pickle\n",
    "import seaborn as sns\n",
    "pal = sns.color_palette(\"Set1\", 10)\n",
    "sns.set(palette=pal, style='whitegrid', \n",
    "       font_scale=1.1, font='serif')\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>post_message</th>\n",
       "      <th>post_id</th>\n",
       "      <th>date_post</th>\n",
       "      <th>post_type</th>\n",
       "      <th>post_link</th>\n",
       "      <th>post_picture</th>\n",
       "      <th>post_contentlink</th>\n",
       "      <th>link_domain</th>\n",
       "      <th>shares</th>\n",
       "      <th>likes_count_fb</th>\n",
       "      <th>comments_count_fb</th>\n",
       "      <th>comment_message</th>\n",
       "      <th>comment_likes</th>\n",
       "      <th>date_comment</th>\n",
       "      <th>comment_id</th>\n",
       "      <th>comment_is_reply</th>\n",
       "      <th>comment_position</th>\n",
       "      <th>user_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PEGIDA updated their cover photo.</td>\n",
       "      <td>790669100971515_836275919744166</td>\n",
       "      <td>2014-12-29 09:22:57</td>\n",
       "      <td>photo</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/posts/836275919744166</td>\n",
       "      <td>https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&amp;oe=56A00BB2&amp;__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3</td>\n",
       "      <td>facebook.com</td>\n",
       "      <td>885</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>👍</td>\n",
       "      <td>1</td>\n",
       "      <td>2014-12-29 09:23:43</td>\n",
       "      <td>836275903077501_836276183077473</td>\n",
       "      <td>0</td>\n",
       "      <td>131_1</td>\n",
       "      <td>750d1a75ad399eaffa01d0802a25ad3c497fd181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PEGIDA updated their cover photo.</td>\n",
       "      <td>790669100971515_836275919744166</td>\n",
       "      <td>2014-12-29 09:22:57</td>\n",
       "      <td>photo</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/posts/836275919744166</td>\n",
       "      <td>https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&amp;oe=56A00BB2&amp;__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3</td>\n",
       "      <td>facebook.com</td>\n",
       "      <td>885</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>Lasst die Erde beben !!</td>\n",
       "      <td>17</td>\n",
       "      <td>2014-12-29 09:24:20</td>\n",
       "      <td>836275903077501_836276326410792</td>\n",
       "      <td>0</td>\n",
       "      <td>131_3</td>\n",
       "      <td>d715e7a5ea13d85e6fb7d47ff5bc41d0050a1bc0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PEGIDA updated their cover photo.</td>\n",
       "      <td>790669100971515_836275919744166</td>\n",
       "      <td>2014-12-29 09:22:57</td>\n",
       "      <td>photo</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/posts/836275919744166</td>\n",
       "      <td>https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&amp;oe=56A00BB2&amp;__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3</td>\n",
       "      <td>https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3</td>\n",
       "      <td>facebook.com</td>\n",
       "      <td>885</td>\n",
       "      <td>4208</td>\n",
       "      <td>2547</td>\n",
       "      <td>Hut ab,nur so gehts.Ich wünsche Euch allen viel Erfolg &amp; einen guten Rutscht ins Jahr 2015</td>\n",
       "      <td>13</td>\n",
       "      <td>2014-12-29 09:26:12</td>\n",
       "      <td>836275903077501_836277516410673</td>\n",
       "      <td>0</td>\n",
       "      <td>131_4</td>\n",
       "      <td>de3371443aef9951c2288513de175a8b5feb7b3b</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        post_message                          post_id  \\\n",
       "0  PEGIDA updated their cover photo.  790669100971515_836275919744166   \n",
       "1  PEGIDA updated their cover photo.  790669100971515_836275919744166   \n",
       "2  PEGIDA updated their cover photo.  790669100971515_836275919744166   \n",
       "\n",
       "            date_post post_type  \\\n",
       "0 2014-12-29 09:22:57     photo   \n",
       "1 2014-12-29 09:22:57     photo   \n",
       "2 2014-12-29 09:22:57     photo   \n",
       "\n",
       "                                                        post_link  \\\n",
       "0  https://www.facebook.com/pegidaevdresden/posts/836275919744166   \n",
       "1  https://www.facebook.com/pegidaevdresden/posts/836275919744166   \n",
       "2  https://www.facebook.com/pegidaevdresden/posts/836275919744166   \n",
       "\n",
       "                                                                                                                                                                                                                   post_picture  \\\n",
       "0  https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&oe=56A00BB2&__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3   \n",
       "1  https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&oe=56A00BB2&__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3   \n",
       "2  https://fbcdn-photos-g-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/s130x130/1480490_836275903077501_255733933142326293_n.jpg?oh=9874686e3780c41310678f25768f02a8&oe=56A00BB2&__gda__=1452611089_fcc9c22a1bfd4af2ac50968d4467dae3   \n",
       "\n",
       "                                                                                                       post_contentlink  \\\n",
       "0  https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3   \n",
       "1  https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3   \n",
       "2  https://www.facebook.com/pegidaevdresden/photos/a.790680730970352.1073741826.790669100971515/836275903077501/?type=3   \n",
       "\n",
       "    link_domain  shares  likes_count_fb  comments_count_fb  \\\n",
       "0  facebook.com     885            4208               2547   \n",
       "1  facebook.com     885            4208               2547   \n",
       "2  facebook.com     885            4208               2547   \n",
       "\n",
       "                                                                              comment_message  \\\n",
       "0                                                                                           👍   \n",
       "1                                                                     Lasst die Erde beben !!   \n",
       "2  Hut ab,nur so gehts.Ich wünsche Euch allen viel Erfolg & einen guten Rutscht ins Jahr 2015   \n",
       "\n",
       "   comment_likes        date_comment                       comment_id  \\\n",
       "0              1 2014-12-29 09:23:43  836275903077501_836276183077473   \n",
       "1             17 2014-12-29 09:24:20  836275903077501_836276326410792   \n",
       "2             13 2014-12-29 09:26:12  836275903077501_836277516410673   \n",
       "\n",
       "   comment_is_reply comment_position                                   user_id  \n",
       "0                 0            131_1  750d1a75ad399eaffa01d0802a25ad3c497fd181  \n",
       "1                 0            131_3  d715e7a5ea13d85e6fb7d47ff5bc41d0050a1bc0  \n",
       "2                 0            131_4  de3371443aef9951c2288513de175a8b5feb7b3b  "
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comments = pd.read_excel('data/pegida_fb_merged_Dec14-May16.xlsx') # encoding utf8\n",
    "\n",
    "comments.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>counts</th>\n",
       "      <th>relcounts</th>\n",
       "      <th>cumcounts</th>\n",
       "      <th>relcum</th>\n",
       "      <th>reluser</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>f2a8223a1bbed529651d5c1ff81c8b02bdf4363e</th>\n",
       "      <td>1</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.00001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a97d9d64b76d6c7003bef567c99b4615d2725241</th>\n",
       "      <td>1</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.00002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          counts  relcounts  cumcounts  \\\n",
       "user_id                                                                  \n",
       "f2a8223a1bbed529651d5c1ff81c8b02bdf4363e       1   0.000001          1   \n",
       "a97d9d64b76d6c7003bef567c99b4615d2725241       1   0.000001          2   \n",
       "\n",
       "                                            relcum  reluser  \n",
       "user_id                                                      \n",
       "f2a8223a1bbed529651d5c1ff81c8b02bdf4363e  0.000001  0.00001  \n",
       "a97d9d64b76d6c7003bef567c99b4615d2725241  0.000002  0.00002  "
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped = comments.groupby('user_id')\n",
    "size = grouped.size().to_frame('counts').sort_values('counts')\n",
    "size['relcounts'] = size.counts / len(comments)\n",
    "size['cumcounts'] = size.counts.cumsum()\n",
    "size['relcum'] = size.cumcounts / len(comments)\n",
    "size['reluser'] = 1\n",
    "size['reluser'] = size.reluser.cumsum() / len(size)\n",
    "size.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>counts</th>\n",
       "      <th>relcounts</th>\n",
       "      <th>cumcounts</th>\n",
       "      <th>relcum</th>\n",
       "      <th>reluser</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7e359216fede59332fabbb810ee7d1218b3a07d1</th>\n",
       "      <td>1512</td>\n",
       "      <td>0.001621</td>\n",
       "      <td>909642</td>\n",
       "      <td>0.975117</td>\n",
       "      <td>0.999911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ae4b902f4b5620352ef40fdd9868bdb84d636fdd</th>\n",
       "      <td>1597</td>\n",
       "      <td>0.001712</td>\n",
       "      <td>911239</td>\n",
       "      <td>0.976829</td>\n",
       "      <td>0.999921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4594b86870233901e4e9aa086279dfd366fe5ee0</th>\n",
       "      <td>1651</td>\n",
       "      <td>0.001770</td>\n",
       "      <td>912890</td>\n",
       "      <td>0.978599</td>\n",
       "      <td>0.999931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c44114ffb73318cd4cc7f683a391b6dbb9be7129</th>\n",
       "      <td>1723</td>\n",
       "      <td>0.001847</td>\n",
       "      <td>914613</td>\n",
       "      <td>0.980446</td>\n",
       "      <td>0.999941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15b7f8b81411045f3022f96c2ddbaabc04da4b5f</th>\n",
       "      <td>1833</td>\n",
       "      <td>0.001965</td>\n",
       "      <td>916446</td>\n",
       "      <td>0.982411</td>\n",
       "      <td>0.999950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e79c8f4acf98549ed212a5108583725e680a5374</th>\n",
       "      <td>1915</td>\n",
       "      <td>0.002053</td>\n",
       "      <td>918361</td>\n",
       "      <td>0.984464</td>\n",
       "      <td>0.999960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8d42d01e8551a380c12a48249825db155729ac14</th>\n",
       "      <td>1917</td>\n",
       "      <td>0.002055</td>\n",
       "      <td>920278</td>\n",
       "      <td>0.986519</td>\n",
       "      <td>0.999970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d3b0b6f65ef4e4d53ba56e9aca7d90b201b958ce</th>\n",
       "      <td>1999</td>\n",
       "      <td>0.002143</td>\n",
       "      <td>922277</td>\n",
       "      <td>0.988662</td>\n",
       "      <td>0.999980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4bbf6251cd91985390a6b98336cd6b4030a7011a</th>\n",
       "      <td>2871</td>\n",
       "      <td>0.003078</td>\n",
       "      <td>925148</td>\n",
       "      <td>0.991739</td>\n",
       "      <td>0.999990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6286c39c170249860ec1fcb370b91a90a18f577d</th>\n",
       "      <td>7706</td>\n",
       "      <td>0.008261</td>\n",
       "      <td>932854</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          counts  relcounts  cumcounts  \\\n",
       "user_id                                                                  \n",
       "7e359216fede59332fabbb810ee7d1218b3a07d1    1512   0.001621     909642   \n",
       "ae4b902f4b5620352ef40fdd9868bdb84d636fdd    1597   0.001712     911239   \n",
       "4594b86870233901e4e9aa086279dfd366fe5ee0    1651   0.001770     912890   \n",
       "c44114ffb73318cd4cc7f683a391b6dbb9be7129    1723   0.001847     914613   \n",
       "15b7f8b81411045f3022f96c2ddbaabc04da4b5f    1833   0.001965     916446   \n",
       "e79c8f4acf98549ed212a5108583725e680a5374    1915   0.002053     918361   \n",
       "8d42d01e8551a380c12a48249825db155729ac14    1917   0.002055     920278   \n",
       "d3b0b6f65ef4e4d53ba56e9aca7d90b201b958ce    1999   0.002143     922277   \n",
       "4bbf6251cd91985390a6b98336cd6b4030a7011a    2871   0.003078     925148   \n",
       "6286c39c170249860ec1fcb370b91a90a18f577d    7706   0.008261     932854   \n",
       "\n",
       "                                            relcum   reluser  \n",
       "user_id                                                       \n",
       "7e359216fede59332fabbb810ee7d1218b3a07d1  0.975117  0.999911  \n",
       "ae4b902f4b5620352ef40fdd9868bdb84d636fdd  0.976829  0.999921  \n",
       "4594b86870233901e4e9aa086279dfd366fe5ee0  0.978599  0.999931  \n",
       "c44114ffb73318cd4cc7f683a391b6dbb9be7129  0.980446  0.999941  \n",
       "15b7f8b81411045f3022f96c2ddbaabc04da4b5f  0.982411  0.999950  \n",
       "e79c8f4acf98549ed212a5108583725e680a5374  0.984464  0.999960  \n",
       "8d42d01e8551a380c12a48249825db155729ac14  0.986519  0.999970  \n",
       "d3b0b6f65ef4e4d53ba56e9aca7d90b201b958ce  0.988662  0.999980  \n",
       "4bbf6251cd91985390a6b98336cd6b4030a7011a  0.991739  0.999990  \n",
       "6286c39c170249860ec1fcb370b91a90a18f577d  1.000000  1.000000  "
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "size.tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usershare</th>\n",
       "      <th>contentshare</th>\n",
       "      <th>usershare_p</th>\n",
       "      <th>contentshare_p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.1000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>10.0%</td>\n",
       "      <td>1.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.2500</td>\n",
       "      <td>0.03</td>\n",
       "      <td>25.0%</td>\n",
       "      <td>3.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.5000</td>\n",
       "      <td>0.06</td>\n",
       "      <td>50.0%</td>\n",
       "      <td>6.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.7500</td>\n",
       "      <td>0.15</td>\n",
       "      <td>75.0%</td>\n",
       "      <td>15.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.9000</td>\n",
       "      <td>0.30</td>\n",
       "      <td>90.0%</td>\n",
       "      <td>30.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.9500</td>\n",
       "      <td>0.42</td>\n",
       "      <td>95.0%</td>\n",
       "      <td>42.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.9900</td>\n",
       "      <td>0.68</td>\n",
       "      <td>99.0%</td>\n",
       "      <td>68.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.9990</td>\n",
       "      <td>0.90</td>\n",
       "      <td>99.9%</td>\n",
       "      <td>90.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.9999</td>\n",
       "      <td>0.97</td>\n",
       "      <td>99.99%</td>\n",
       "      <td>97.0%</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   usershare  contentshare usershare_p contentshare_p\n",
       "0     0.1000          0.01       10.0%           1.0%\n",
       "1     0.2500          0.03       25.0%           3.0%\n",
       "2     0.5000          0.06       50.0%           6.0%\n",
       "3     0.7500          0.15       75.0%          15.0%\n",
       "4     0.9000          0.30       90.0%          30.0%\n",
       "5     0.9500          0.42       95.0%          42.0%\n",
       "6     0.9900          0.68       99.0%          68.0%\n",
       "7     0.9990          0.90       99.9%          90.0%\n",
       "8     0.9999          0.97      99.99%          97.0%"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "percs = [0.1, 0.25,0.5,0.75, 0.9,0.95,0.99, 0.999,0.9999]\n",
    "perc_list = []\n",
    "for i in percs:\n",
    "    perc_list.append([i, size.loc[size.reluser >= i].head(1)['relcum'].values[0]])\n",
    "shares = pd.DataFrame(perc_list, columns=[\"usershare\", 'contentshare'])\n",
    "shares['contentshare'] =  shares['contentshare'].round(2)\n",
    "shares['usershare_p'] =  (shares.usershare * 100).astype(str) + \"%\"\n",
    "shares['contentshare_p'] =  (shares.contentshare * 100).astype(str) + \"%\"\n",
    "shares"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7, 5))\n",
    "g = sns.pointplot(x=\"usershare\", y=\"contentshare\", data=shares,\n",
    "                 ci=False, scale=0.8,\n",
    "                 color = pal[3])\n",
    "plt.ylim(0.0,1.0)\n",
    "vals = [0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9, 1.0]\n",
    "plt.yticks(vals, [str(i * 100) + '%' for i in vals])\n",
    "g.set_xticklabels(shares.usershare_p, rotation=30)\n",
    "g.set_xlabel('Cumulative Users', fontsize=12)\n",
    "g.set_ylabel('Cumulative Comments', fontsize=12)\n",
    "g.xaxis.grid(True)\n",
    "g.yaxis.grid(True)\n",
    "\n",
    "plt.tight_layout(pad=3)\n",
    "sns.despine()\n",
    "plt.savefig('pegida_suppfig_usercomments.jpeg', format='jpeg', dpi=300)\n",
    "plt.show() "
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}
